Own the Press - Distribution Is Power

Within fifty years of Gutenberg's workshop going live, the institutions of Europe had worked out the only fact about printing that ever mattered: the danger was never the book. It was the press.

You can burn a book; another copy exists. But control the press — license it, tax it, register it, count it — and you control everything not yet written. England's Licensing of the Press Act made it law: no printing without a state-approved license, no presses outside the approved cities, and the Stationers' Company authorized to search any shop where an unregistered machine might be hiding. The censors understood their business precisely. Ideas are downstream of distribution. Whoever owns the channel owns the conversation, and whoever owns the conversation, in the long run, owns the consensus.

Five centuries later the machinery has changed and the architecture has not. This is the piece where I explain why a person who writes seriously in 2026 should think like a seventeenth-century printer — and what it looks like when one finally owns the press.

The license came back wearing better clothes

Nobody stamps a royal seal on your publishing license today. Instead you receive it for free, instantly, from a platform — and almost nobody reads the terms of what they've been issued.

The modern license works by ranking instead of registration. You may print whatever you like; the question is whether anyone is permitted to see it. An algorithm sits between you and every reader you've ever earned, and it adjudicates — continuously, opaquely, revocably — how far each sentence travels. Reach is granted and withdrawn like a privilege, because that is what it is. The seventeenth-century printer at least knew where he stood: the license named its terms, and losing it was a legible event. The modern writer's license degrades silently — a distribution curve flattens, a format falls from favor, a topic quietly stops traveling — and there is no notice, no hearing, and no appeal that reaches a human.

Notice, too, the asymmetry of the paperwork. You agreed to the platform's terms — thousands of words, revisable at will, written by their counsel for their protection. Nobody agreed to yours. You have no terms. There is no document anywhere in which the platform undertakes to keep showing your work to the people who asked to see it. The entire relationship is a license flowing one direction, dressed in the vocabulary of partnership — creator, community, your audience — words that describe everything except the legal reality, which is tenancy.

I want to be precise rather than dramatic, because the mechanism doesn't need drama. The platforms are not villains; they are landlords, and they behave exactly as the incentives of landlords predict. The feed optimizes for the platform's revenue, not your reach. Moderation optimizes for the platform's legal exposure, not your nuance. None of it is aimed at you personally — which is exactly why none of it can be negotiated with. You cannot appeal to an incentive structure. You can only stop depending on it.

The same mistake, without a passport

Old-line sovereignty thinking — the flag-theory tradition — starts from one master principle: never let a single system hold the power to erase you. People who handle money seriously have internalized this for decades. They would never keep every asset in one bank, in one currency, under one jurisdiction, because a single policy change shouldn't be able to zero out a life.

Then the same people pour ten years of writing, an audience, and a reputation into a single platform account — one company, one terms-of-service, one suspension queue — and call it a brand. It is the same mistake. It just hasn't got a passport.

This is, in the old taxonomy, a digital-security problem — the flag you plant yourself, with no government's cooperation and no institution's permission. A domain you register. A server you control. A database you can copy. Software you can move. The remarkable thing about this particular flag is its price: the cheapest in the whole portfolio, a few hundred dollars a year and some competence — against which the alternative is renting your public existence from a counterparty who can repossess it. (And to say it once cleanly, since sovereignty talk attracts the other kind: everything here is structure on the legal side of every line — ownership, not evasion of anything.)

The objection arrives on schedule: self-hosting is harder than signing up. True, on day one. But "easy to start" and "safe to depend on" are different axes, and confusing them is how a decade of work ends up on land you don't own. Hard-to-start is a one-time cost. Impossible-to-depend-on is a permanent one.

What a press actually is, in 2026

Here I should declare an interest, openly: I didn't just reach these conclusions — I acted on them, as an engineer. I build and run my own publishing infrastructure, and the technical site where I show that work is rist.sh; this site is where I show the thinking. What follows isn't a product tour. It's an answer to a more useful question: when an engineer sits down and asks what "owning a press" actually requires, what does the machine have to do?

It has to hold the content in a form you own. Not one undifferentiated blob of posts — typed content: articles, project case studies, changelogs, events, books, courses, each a shape the system understands, stored in a database you can dump, move, and restore on any machine on earth. Export as a first-class function, not a buried courtesy.

It has to decide who sees what — on your terms, enforced by you. This was the part that surprised me most when I designed it: access is the press's real power. The machine I run resolves visibility along three independent axes — how trusted a reader is, which groups they belong to, and a stack of per-post locks: passwords, secret links, scheduled release, early access, expiry, geographic rules. The axes compose, and the strictest rule always wins. Drawn on paper, it looks like a map of concentric audiences — the public at the edge, registered readers deeper, inner circles deeper still — with you deciding, structurally rather than wishfully, which work lives at which depth. The seventeenth-century press could only print. A modern press gates — it is a printing house and a private reading room in one machine.

It has to separate the voice from the account. Bylines on my press are personas — public identities decoupled from logins, each with its own name and archive. A publication is a stage, and a stage needs more than one mask that can speak; any serious press supports that as architecture, not as a hack.

It has to publish without you standing at the lever. Software can file its own reports onto my press through a typed publishing interface — but software earns trust the way a junior writer does, in stages: first everything lands as a draft, then with one-tap approval, then autonomously, with quotas at every stage. The press extends to machines without surrendering to them.

It has to treat addresses as commitments. A reader's bookmark, a citation in someone else's essay, a search engine's accumulated trust in a page — all of it is keyed to a URL. On my press, an address, once published, never silently changes; when one must move, the old address forwards to the new one permanently. Rented platforms reorganize their URL schemes whenever it suits them, and every inbound link built over years quietly dies. A press that breaks its own addresses is burning its own archive.

And it has to fan out instead of funneling in. Every piece flows outward to feeds, syndication formats, and the platforms — which inverts the power relation. The platforms become spokes: places my work visits, not places it lives. The press is the hub. Nobody ranks me out of my own building.

None of this is hypothetical — both of my deployments are live in production, and the engineering detail is documented, piece by piece, on the technical site. The point here is the shape: that is what a press is now. A machine that holds, gates, signs, automates, and distributes — under one owner, on owned hardware.

Samizdat ran on worse

When the licensing regimes of the twentieth century were at their most total, the counter-move was almost comically physical: carbon paper. Soviet samizdat circulated entire banned literatures through typewriters and onion-skin copies, four or five legible sheets at a time, hand to hand. People risked years of their lives for distribution bandwidth that a modern phone exceeds by twelve orders of magnitude.

I think about that asymmetry often, and it cures all self-pity about the difficulty of self-hosting. The dissidents of that era would have wept at what is legally available to anyone today for the price of a dinner: a global press, instant distribution, cryptographic signatures, and no Stationers' Company with the right to search the premises. The machinery of independent publication has never — never, in the entire history of the written word — been this cheap, this legal, and this powerful. The only thing rarer than it should be is the decision to use it.

The press is a position, not a purchase

Strip away the engineering and the history, and what's left is the strategic point, which is the one I actually care about.

Owning your press changes your position — in the precise, almost chess-free sense of where you stand and what can be done to you. The writer on a platform argues from rented ground: every sharp sentence is weighed against an invisible moderation curve, every audience relationship is intermediated by a counterparty with veto power. The writer on owned ground argues from a different posture entirely. Not louder — sturdier. You write what you judge true and useful, and the only standard you answer to is your own, enforced by infrastructure rather than promised by policy.

That position compounds. Every essay added to an owned archive appreciates there permanently; every essay poured into a feed depreciates in hours. Every reader who arrives at your domain is a relationship you keep; every follower on a platform is a relationship you're permitted to borrow. Run those two curves for a decade and they describe two different lives.

The first move costs an afternoon: a domain with your name on it, anything self-hosted behind it, and the decision — which is the actual asset — that your work will accumulate on land you own. Move that one flag and you'll feel the position shift under your feet. The rest of the portfolio can wait.

A press, though, is machinery — and machinery says nothing about the hand that built it, or why a person trained to read systems would spend years constructing one instead of renting. That's the next essay: the mind behind the machine.

---

Summary

From Gutenberg to the feed, whoever owns the channel owns the conversation. The case for owning a press — and what one actually is now.

The License Returned. Modern platforms reissue the seventeenth-century printing license as algorithmic reach — granted free, adjudicated opaquely, revoked silently, no human appeal.

The Same Mistake. Nobody serious keeps all assets in one bank, yet writers keep a decade of work in one account. Platform risk is sovereign risk without the passport.

What a Press Is Now. A machine that holds typed content you can move, gates audiences along composable axes, separates voices from accounts, lets software earn trust in stages, and fans out to platforms as spokes, not homes.

Samizdat Ran on Worse. Dissidents risked years for carbon-copy bandwidth; a legal, global press now costs a dinner. The rare part is the decision.

Position, Not Purchase. Owned archives appreciate; fed content depreciates in hours. The press changes where you stand.